[Wikimedia-l] Fundraising mysteries

[James Salsman]

If October 4th and/or 22nd had large donations because of one-time events
instead of regular donation appeal changes, why are they both bracketed by
vastly abnormally successful previous and subsequent days?


On Monday, November 25, 2013, James Salsman wrote:

> P.S. As the referenced attachment doesn't make it through to the
> archives or digests, there is a copy of the fundraising data graph at:
> http://i.imgur.com/MkXIW4J.png
>
>
> On Mon, Nov 25, 2013 at 7:51 PM, James Salsman <jsalsman at gmail.com> wrote:
> > Hi Megan,
> >
> > Per the attached graph of the
> > https://frdata.wikimedia.org/yeardata-day-vs-sum.csv
> > data, your announced October 4th fundraising test on 100% of anonymous
> > users was anticlimactic. But what the heck did you do on September 2nd
> > and October 22nd, and would you please do that every day? Even if it
> > falls off at the same rate as the July test, that still means you
> > could produce an endowment sufficient to do away with fundraising at
> > current spending levels in less than eight months.
> >
> > Best regards,
> > James Salsman
>